Desire Paths

Our designs, our art, our gardens, subways and startups, none of them reach their potential unless they have users. We talked about stakeholders a few weeks ago, and no stakeholder is more key than the one that interacts directly with the system. Our users; customers, subscribers, clients, members… they form a community, one that includes us but very rarely are we the center of that community. The fact that we’ve ‘got’ them doesn’t mean they will act as we would or appreciate what we deliver, or interact in the way that we intend.

For these reasons, getting our outward design, the boundary of interface between what we build and those that use it, correct is possibly the most critical piece of our system. At the same time, for the same reasons, it’s the piece we are most likely to get wrong. Recall your favorite municipal park, the manicured lawns interspersed with well-laid paths that lead between utilities and attractions. Someone, probably a team trained in Urban Design, architecture and lansdcaping, spent effort laying out those paths multiple times, over a careful and iterative process. They took great effort to consider traffic expectations, human ergonomics, aesthetics, even economics and social theory.

But now really think about your mind picture. How often are there little additional paths, forks that lead off, not built by a planning committee, but just dirt tracks in the professionally laid sod that lead off the main path?

Those dirt tracks, called desire lines are caused by disgruntled users - used by enough of them that they form their own permanent paths, an emergent design laid by a committee of users. That design of actual usage lays over the design of intended use in everything we build, every app we program, every legal system we construct. When the delta V of our intended usage strays too far from the actual usage, we create discord in our users. People stop coming to events, they break laws, they cancel subscriptions and they begin to think of your design as outdated and out of touch.

Crosswalks are one example of a necessary security feature of public roads. No matter how well placed and well painted, people are far too likely to run against the light, putting themselves at risk. By introducing a crosswalk button to an intersection, we return a sense of control to the user - people are far more likely to wait for a light change if there is a button available, even though the great majority of those buttons aren’t hooked up to anything. Most municipal crosswalks run on timers to improve the flow of vehicle and pedestrian traffic, the buttons are simply placebos installed to keep people safer.

There are four major strategies to cope with user discord: education, regulation, flexibility and iteration.

Education is one of the least effective, but probably the most frequently employed. Improving signage, writing manuals, community outreach, and tutorials have their place, but in each case, it’s a matter of educating someone to change a natural behavior that they probably have no interest in changing. They aren’t committed to the process: probably fairly, since it’s your design that isn’t working. Learning something new is seen as a chore by many people, and learning the “better” way you envision them behaving seems both wasteful and rude. Writing documentation is important, but expecting the end user to absorb it before using your software, train system or cooperative store is a sure to create user discord.

Failing education strategies are often bolstered with regulation. If people keep making their own paths, some parks will put in bushes, fences, art pieces, physical barriers to keep them on the approved path. To control vehicles capable of dangerous speeds driven by people who simultaneously misunderstand basic physics, the importance of safety and the relative importance of getting someplace as quickly as possible, we have the highway patrol, ready to punish those who take advantage of that system. Entering a phone number or postal code in most web forms either requires you to type in a specific format, or the form returns an error after submission. When thousands of stock brokers panic and threaten to crash the fragile stock market, automatic brakes kick in preventing motion in the system.

Flexibility is the most comfortable of these strategies. By giving up control, we can improve the user experience, although it often comes at a greater cost in materials and other resources. Allowing the user to act as they with requires better planning, more upkeep and also requires the designer to relinquish some goals, something that can be hard on the ego. The self-made paths in parks mentioned above are an example of a flexible system, providing an outlet for users that disagree with the designed layout. But those paths will be more prone to erosion, creating muddy puddles when it rains, and washing out completely from time to time. It decreases the intended aesthetic value of the park, and may provide a path that puts people in greater danger, out of the protection of street lights. Providing a field on a web form that accepts freeform phone numbers means some people will enter 7 digits, others 10 and some may type their email. Somewhere along the chain, when that number needs to be used, it requires additional dexterity on the programmer or caller to deal with the different formats that will happen. It will mean some people will not be contactable and might cause losses in services or missed packages and opportunities. The amazing time management software Freckle lets me create a new project from anywhere in the system, just by typing an unknown name - not only saving my time, but also making me a super fan.

Iteration is the most effective strategy, but also the most expensive. By iterating our designs, we watch how users interface with them and then modify the design to cause less discord. Incremental changes could include formalizing the paths that users create and rerouting pavement or cobbles. It might mean rezoning and relocating a shopping district to improve traffic, or planting more raspberries in Zone 0 instead of Zone 1, so they are closer to the house. It could mean a complete redesign of the signup process on your SAAS app, to eliminate cart abandonment, or it may mean changing the tense of a few words in the process to help people understand the order they are expected to proceed.

The ways users relate to our designs are complicated concepts, too often left to designers with little experience or too little care to prevent discord. Amazing user experiences make amazing customers, members and patrons, and we have to consider improving them to be a worthy investment in all of our designs.


New here? Reading what is permaculture” will help you catch up quickly.

Image by Duncan Rawlinson